Hypothesis testing Flashcards Preview

Stats > Hypothesis testing > Flashcards

Flashcards in Hypothesis testing Deck (18)
Loading flashcards...
1

What are parametric data?

data that are normally distrubuted

2

What are 3 types of hypotheses you can test?

1. Single sample (comparing a sample to a known value e.g. average weight of woman in USA)

2. two independent samples (comparing two experimental conditions and different participants were assigned to each condition)
- two groups of subjects in two different conditions (experimental and control)

3. two dependent samples (two experimental conditions and the same participants took part in both conditions of the experiment)
- same group of subjects in the conditions

3

What is the null hypothesis?

he position that any difference is simply due to sampling error (chance variation)

4

What is the alternate hypothesis?

the position that any difference is not due to sampling error but rather due to the effect of the experimental manipulation

5

What are the 8 steps of hypothesis testing?

1. choose a hypothesis to test
2. choose an alternate hypothesis
3. choose a significance level
4. determine the distribution
5. determine the critical region
6. establish a rejection rule
7. calculate the statistic
8. draw the appropriate conclusion

6

What is alpha

he probability of making a Type I error

7

what is beta

the probability of making a Type II error

8

what is power

the probability of correctly rejecting the null hypothesis when the null hypothesis is false (i.e. the probability of NOT making a type 2 error)

Probability that a given test will find an effect assuming that one exists in the population

9

what is a type 2 error

Incorrectly failing to reject the null hypothesis when the null hypothesis is false

10

what is a type 2 error

Incorrectly failing to reject the null hypothesis when the null hypothesis is false

11

What factors influence alpha?

the researcher (based on the consequences of committing a type 1 error)

12

What are the factors that affect beta? (6)

1. alpha
2. the alternative hypothesis
3. the size of the samples (large sample = more power)
4. the real difference between the two means (the magnitude of the true difference i.e. effect size)
5. parametric vs non-parametric tests (parametric have more power)
6. independent- vs dependent-groups designs

13

how do you calculate the probability of rejecting the null hypothesis? (power)

Need to base this on comparing 2 distributions
- the distribution you would get if the null hypothesis was true vs if the null is false

14

How does independent- vs dependent-groups designs impact power?

By impacting amount of variability
-> dependent group designs have fewer sources of variability and therefor more power to detect a treatment effect

15

What are the sources of variability in an dependent groups (within groups) t test?

1. Experimental error
2. Treatment effect

16

What are the sources of variability in an independent groups (between groups) t test?

1. Experimental error
2. Individual differences
3. Treatment effect

17

What is experimental error

1. fluctuations in external conditions (e.g. room temperature, lighting)
2. temporary variations in the subjects themselves (e.g. distracted, tired)

18

What are individual differences

Stable differences between individuals (e.g. intelligence, personality factors)